Hi,
It has been a long time since I wrote a sort or a search.
I know that the bubble sort is a dog compared to a comb sort.
I know that a sequential search is a dog compared to a binary search.
I am just wondering if the name searches in the global variable DLL are
smart enough to sort new named items that are first found back into the
tables via a fast comb sort and when it does its search for each new
submission
item if it uses something fast such as the binary search to find items
in sorted data.
I don't know what hashing means.
Maybe things are better now than comb and binary ????
The bottom line questions is this.
It it worth working out a map for all the items in the global variables
using pointers (numbers in arrays) or are the methods
of finding data in the DLL routines via string descripters so fast that
it is not
worth the time to work out the exact numbers.
Machines may be fast but every bit helps.
Thanks,
John.
|